Quote:
Originally Posted by 4bama 
Speedtest.net uses a Flash based byte counter and it's notoriously inaccurate for measuring true download and upload speeds.
The one speed test site that most serious testers use is at testmy.net, which uses only your browser for the test and allows the user to run separate download and upload speed tests, and allows the user to select a fixed file size for those tests...never use the "SmartTest" option if you want true accuracy in your speeds.
If you go to the web site below, run two separate tests..click the "Testmy Download" button and when that page loads select a 25mB or 50mB file from the drop-down box as your test size and let it run...
Then, go to the "Testmy Upload" option and select the 17 or 23mB file from the drop-down box as your test size and let it run..
You can post those results...if you run these tests early in the morning when fewer customers are competing for bandwidth on your ISP/Internet connection, then run them again during the peak-use hours of 7-10pm you will have a fair picture of how congestion effects your speed.
http://testmy.net/

Speedtest.net uses a Flash based byte counter and it's notoriously inaccurate for measuring true download and upload speeds.
The one speed test site that most serious testers use is at testmy.net, which uses only your browser for the test and allows the user to run separate download and upload speed tests, and allows the user to select a fixed file size for those tests...never use the "SmartTest" option if you want true accuracy in your speeds.
If you go to the web site below, run two separate tests..click the "Testmy Download" button and when that page loads select a 25mB or 50mB file from the drop-down box as your test size and let it run...
Then, go to the "Testmy Upload" option and select the 17 or 23mB file from the drop-down box as your test size and let it run..
You can post those results...if you run these tests early in the morning when fewer customers are competing for bandwidth on your ISP/Internet connection, then run them again during the peak-use hours of 7-10pm you will have a fair picture of how congestion effects your speed.
http://testmy.net/
Unfortunately the post above is factually incorrect (but is the mantra that TMN is so well known for). The facts are that both STN and TMN use what's known in the industry as http based testing. In the case of STN what you the user sees as flashy eye candy is NOT the actual test but in fact is a graphical representation of the test progression designed to keep your attention. In reality both use the same general method for testing behind the scenes which amounts to downloading a highly compressed file of a specific size and then computing the result based on the time it takes to download that file.
While this testing method is fine for short hauls it becomes increasingly inaccurate as the distance from the test server increases. BOTH of these tests also fail to properly measure the key component of broadband connections that matters far more than speed and that is datarate consistency aka QoS. You can have a broadband connection claiming to offer you 100mbps but if it fluctuates wildly it's of no more value than it's lowest point of fluctuation especially as it applies to single socket applications like streaming and gaming.
Currently there is only one testing platform that accurately and consistently tests speeds AND QoS and especially over longer distances available to consumers using a TCP socket based method and that is the platform offered by any site that uses Visualware software. It's also the very same platform used by the majors like Cisco, Juniper, AT&T you name it behind the scenes.
As for the HTTP based tests like STN and TMN the only industry accepted test would be that of STN (aka Ookla) but again I caution you to test to your closest server only and you must understand that its only giving you part of the broadband connection equation. You are still missing the most inportant test which is datarate consistency.
TMN does a great job of talking the talk and their schpeal is certainly convincing enough....but don't bite the hype...because that's all it is especially when they compare themselves to STN...for all intents they are one in the same tests with different visual packaging and nothing more. The huge difference between them is that STN offers countless thousands of server opportunities and their core software by Ookla is industry accepted whereas TMN offers just one main with a few extra donated servers and is not industry accepted.























